Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
Pseudomonas aeruginosa cell membrane protein expression from phenotypically diverse cystic fibrosis isolates demonstrate host specific adaptations Karthik Shantharam Kamath, Dana Pascovici, Anahit Penesyan, Apurv Goel, Vignesh Venkatakrishnan, Ian T. Paulsen, Nicolle H. Packer, and Mark P Molloy J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00058 • Publication Date (Web): 01 Jun 2016 Downloaded from http://pubs.acs.org on June 1, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Pseudomonas aeruginosa cell membrane protein expression from phenotypically diverse cystic fibrosis isolates demonstrate hostspecific adaptations Karthik Shantharam Kamath1, Dana Pascovici2, Anahit Penesyan1, Apurv Goel2, Vignesh Venkatakrishnan1, Ian T Paulsen1, Nicolle H Packer1 and Mark P Molloy1, 2,*. 1. Department of Chemistry and Biomolecular Sciences, Macquarie University, Sydney, Australia. 2. Australian Proteome Analysis Facility, Macquarie University, Sydney, Australia.
*Corresponding author: Mark Molloy, Ph.D. Tel:
+61 2 9850 6218
Fax:
+61 2 9850 6200
Email:
[email protected] Total number of words: 8297
1 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 36
Abstract: Pseudomonas aeruginosa is a Gram-negative, nosocomial, highly adaptable opportunistic pathogen especially prevalent in immuno-compromised cystic fibrosis (CF) patients. The bacterial cell surface proteins are important contributors to virulence, yet the membrane subproteomes of phenotypically diverse P. aeruginosa strains are poorly characterized. We carried out mass spectrometry (MS) based proteome analysis of the membrane proteins of three novel P. aeruginosa strains isolated from the sputum of CF patients and compared protein expression to the widely used laboratory strain, PAO1. Microbes were grown in planktonic growth condition using minimal M9 media and a defined synthetic lung nutrient mimicking medium (SCFM) limited passaging. Two-dimensional LC-MS/MS using iTRAQ labelling enabled quantitative comparisons amongst 3171 and 2442 proteins from the minimal M9 medium and in the SCFM, respectively. The CF isolates showed marked differences in membrane protein expression in comparison to PAO1 including up-regulation of drug resistance proteins (MexY, MexB, MexC) and down-regulation of chemotaxis and aerotaxis proteins (PA1561, PctA, PctB), and motility and adhesion proteins (FliK, FlgE, FliD, PilJ). Phenotypic analysis using adhesion, motility and drug susceptibility assays confirmed the proteomics findings. These results provide evidence for host-specific micro-evolution of P. aeruginosa in the CF lung and shed light on the adaptation strategies used by CF pathogens.
Keywords: Pseudomonas aeruginosa, membrane proteome, mass spectrometry, proteomics, virulence, bacterial evolution and adaptation.
2 ACS Paragon Plus Environment
Page 3 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Introduction: Pseudomonas aeruginosa is a Gram-negative, ubiquitous, opportunistic pathogen with a versatile genome (~5500 genes) which provides capacity to adapt and thrive under diverse conditions1, 2. Multi-drug resistant, biofilm forming and hyper-mutable strains of P. aeruginosa are common etiological agents in conditions including burn wounds, immuno compromised conditions (e.g. HIV) and cystic fibrosis (CF). CF is a recessive, congenital disease caused by a mutation in the cystic fibrosis conductance regulator (CFTR) gene. CFTR mutation leads to impaired movement of electrolytes across the cell resulting in the formation of thick mucus which, in turn, creates a favourable environment for the establishment and growth of microbial pathogens. One of the major causes of CF related morbidity and mortality is infection by P. aeruginosa. Studies show that P. aeruginosa infection accounts for 60-70% of all respiratory tract infections in CF patients by the age of twenty3. Despite host inflammatory defence responses and antibiotic treatment, P. aeruginosa causes chronic infections that persist for the lifetime of patients. Chronic P. aeruginosa infections are associated with adaptation in highly compartmentalized, heterogeneous micro-environments of CF patient lungs3. It is likely that genetic flexibility of P. aeruginosa residing inside lung mucus plugs enables the pathogen to specialize in metabolism as per the availability of surrounding nutrients4. Growth in these conditions also requires combating neighbouring bacterial and fungal pathogens and host defences. P. aeruginosa achieves this by expression of an array of surface-associated and secreted virulence factors including lipopolysaccharides, phenazines, hydrogen cyanide (HCN), exoenzyme S, proteases (LasA, LasB)5. Available nutrients are an important trigger to induce expression of many of these factors. Therefore, it is important to consider bacterial culture conditions in studies aimed at investigating growth in micro-environmental conditions. Proteins of bacterial cell envelope (outer membrane, periplasm, cytoplasmic membrane) often represent the first point of contact with the environment and play pivotal roles in establishing virulence, host cell adhesion, acquiring nutrients, drug efflux, signal transduction, quorum 3 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 36
sensing, and escaping host immune system responses. Despite the important contributions of cell envelope proteins in establishing and maintaining virulence in CF, there are almost no studies focused on characterization of these proteins in P. aeruginosa strains isolated from CF patients, with the majority of our knowledge derived from studies of the cell envelope of the burn wound isolate strain, PAO16-10. We recently reported whole genome sequences and global proteomic characterisation from four CF isolate strains of P. aeruginosa grown in LB medium11. This showed that CF strains had similar sized genomes (ranging 6.1-6.4Mbp) and shared common core features, many of which were different to PAO1. These genomic features of CF strains were reflected in phenotypes including altered biofilm formation, pigmentation and virulence. Global proteomic analysis revealed about 50% unique proteins were shared between CF strains compared to PAO1. CF strains expressed proteins involved in the biosynthesis of several nutrients whereas, these proteins were not required by PAO1 grown in LB medium. These results suggested genetic flexibility of CF strains contribute towards survival and adaptation in CF lungs. To better understand the roles of membrane proteins contributing to the adaptability of P. aeruginosa in the CF lung, we cultured three clinical isolates and PAO1 in a defined synthetic CF medium (SCFM) that mimics nutrient conditions in the CF lung12 and compared membrane proteomes. We contrasted these conditions with growth in the M9-glucose minimal medium through the use of two-dimensional liquid chromatography coupled mass spectrometry (MS) of 4-plex iTRAQ tagged peptides13. We validated the proteomic results using various functional assays to demonstrate the importance of drug resistance proteins, chemotaxis and adhesion in contributing to the specialization of P. aeruginosa adaptation in CF lungs.
4 ACS Paragon Plus Environment
Page 5 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Experimental Procedures: Bacterial strains and growth conditions: P. aeruginosa strains PASS1, PASS2, PASS311 and PAO1 (ATCC 15692) (Supplementary Table S1) were grown on solid Luria-Bertani (LB) agar. For proteomic experiments bacteria were inoculated and grown in the synthetic cystic fibrosis medium (SCFM)12 and M9-minimal salts-glucose (M9) medium directly from the frozen stock solution with minimal passaging. SCFM medium was prepared as per Palmer et al12 with modification. Briefly, all the amino acid stock solutions were prepared in 100mM concentration in deionized water except for tryptophan, aspartate and tyrosine, which were resuspended in 0.2, 0.5 M, and 1.0 M sodium hydroxide respectively. For the final preparation of SCFM, amino acid stocks were added into a buffered base solution per litre of SCFM: 6.5 ml, 0.2 M NaH2PO4 , 6.25 ml, 0.2 M Na2HPO4 , 0.348 ml 1 M KNO3 , 0.122 g NH4Cl, 1.114g KCl, 3.03g NaCl, 10 mM MOPS, 779.6 ml deionized water. Volumes of the amino acid stock solutions added are described in Supplementary Table S2. The pH was adjusted to 6.8 and the medium was filter sterilized. Sterile contents were added to the medium (per litre); 0.606ml, 1 M MgCl2, 1.754 ml, 1 M CaCl2, and 1 ml, 3.6 mM FeSO4.7H2O. The M9-glucose medium was prepared as per vendor’s instructions (Sigma, USA).Overnight bacterial cultures were diluted 1:100 using respective medium and grown in biological triplicates (n=3) at 37°C and constant shaking at 200 rpm till mid-logarithmic phase (Supplementary Figure S1). Cell pellets were collected by centrifugation at 2500g for 10 min at 4°C and washed thrice with phosphate-buffered saline (PBS), pH 7.4. Cell lysis and membrane protein enrichment: Cell pellets were resuspended in 0.5 mL PBS containing benzonase (1:100 v/v, Sigma, USA), cOmplete EDTA-free protease inhibitor cocktail tabletTM (Roche, Germany), and an equal amount of acid-washed glass beads (Sigma, USA). Cells were lysed by bead-beating using a FastPrep FP120 bead-beater apparatus (Savant, USA). After lysis, centrifugation at 2500g for 5 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 36
8 min at 4°C was performed to remove cell debris. Membrane proteins were enriched as described previously14. Briefly, cell lysates were treated with ice-cold 100 mM sodium carbonate solution for one hour at 4°C followed by ultracentrifugation at 115,000g at 4°C. The pellet was washed twice with PBS followed by another round of ultracentrifugation. The resultant membrane pellet was resuspended in 1% SDS in water (w/v). The sample was cleaned by acetone precipitation and resuspended in 250mM triethyl ammonium bicarbonate (TEAB) with 0.05% SDS. Protein concentration was determined using Direct Detect (Merck Millipore, USA) according to the manufacturer’s instructions. In-solution digestion, iTRAQ labelling and mass spectrometry (MS): For in-solution digestion, 60µg of protein from each sample was reduced with 5 mM Tris-(2carboxyethyl) phosphine (TCEP) at 60°C for 1 hour and alkylated with 10 mM methyl methane thiosulfonate (MMTS) at room temperature for 10 min, followed by digestion with trypsin (Promega, USA) in 1:20 ratio (trypsin: protein) at 37°C overnight. Digested proteins were lyophilized and resuspended in 0.5 M TEAB and labelled with iTRAQ 4-plex reagent as per manufacturer’s instructions (Applied Biosystems, USA). Labelling efficiency was checked by MALDI-TOF/TOF. Samples were pooled at equal ratios then dried by vacuum centrifugation. Samples were cleaned using Sep-Pak Light C18 cartridge (Waters, USA) and separated into 15 fractions by strong cation exchange (SCX) chromatography as described in Schilter et al15. Fractions were dried and resuspended in 0.1% TFA, 2% acetonitrile. Samples were analysed using nano-LC−MS/MS coupled to a TripleTOF® 5600 mass spectrometer (ABSciex, USA) with positive nanoflow electrospray analysis and information-dependent acquisition (IDA) mode. In IDA, MS/MS acquisition of the 20 most intense m/z values exceeding a threshold > 150 counts per second (cps) with charge states between 2+ to 4+ were selected for MS/MS analysis following a full MS survey scan and excluded for 20 sec to minimize redundant precursor sampling. Protein identification and quantitation:
6 ACS Paragon Plus Environment
Page 7 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Protein identification and quantification was performed using ProteinPilotTM v4.2 software using the ParagonTM algorithm (ABSciex), thorough ID mode including biological modifications and iTRAQ 4plex quantification. MS/MS data were searched against the database generated by combining the PAO1 protein sequence database retrieved from GenBank (release; December 2012) and in-house generated in-silico translated genome databases of PASS1-3 strains11 (23009 entries) with following settings; Cys Alkylation: MMTS, digestion: trypsin, instrument:5600, quantitative (iTRAQ-4 plex) and bias correction. An UnUsed Protein Score cut-off was set to 2.0 (99% protein identification confidence) and the FDR analysis using PSPEP algorithm was enabled. Peptides were quantified on the basis of the intensity of iTRAQ reporter ions via the Paragon algorithm after background correction and bias correction. Proteins were grouped using the ProGroup16 algorithm included in the ProteinPilot program. Quantitation was filled down for competitors in each protein group17 in order to obtain a realistic overlap between the protein identifications on different iTRAQ runs in the presence of a redundant search database. Identification of differentially expressed proteins: Protein quantities were combined from the three experiments and differentially expressed proteins between the various strains and control were identified. The criterion used required the Stouffer combined p-value to be less than 0.05 and consistent trends of differential expression across the triplicates as described previously by us18. Additionally, fold changes were required to be at least 1.3 fold, proteins had to be quantitated in at least two iTRAQ runs and the trend of the quantification had to be similar across replicates. For generation of GO annotation information for the proteins originating from PASS1-3 strains, the corresponding match in PAO1 locus tag (derived using an in-house nucleotide BLAST) were used. GO annotations were manually retrieved from PseudoCAP database (www.pseudomonas.com) and assembled using Perseus 1.5
20
PloGO
21
and WeGO
19
22
. COG categories for proteins
were derived from NCBI P. aeruginosa genome information 23.
7 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 36
Bioinformatics analysis and predictions: For the prediction of the presence of transmembrane helices (TMH), we utilized TMHMM 2.0 server
24
. The N-terminal signal peptide was predicted using SignalP
25
. Grand average of
hydropathicity (GRAVY) score was calculated (http://www.gravy-calculator.de/). Prediction of the subcellular location was performed using SOSUI-GramN
26
. Hypothetical proteins with no
functional annotation were considered as potential membrane proteins if they contained, at least, one predicted TMH and SOSUI-GramN sub-cellular location prediction annotated as membrane. PAO1 GO annotation information in PseudoCAP database
19
was used for
assignment of proteins into functional categories. Information in VFDB was utilized to assign proteins as virulence factors
27
.Information on P. aeruginosa PAO1 membrane transporters
were manually retrieved from TransportDB database 28. Bacterial sputum binding assay: The sputum sample was collected by non-invasive expectoration from a male patient aged 23 years and diagnosed with CF (∆F508/∆F508 mutation in CFTR gene), at Westmead Hospital, Sydney Human research ethics approval was obtained from the Sydney West Area Health service (HREC/10/WMEAD/180). The patient’s sputum tested positive for the presence of mucoid and non-mucoid P. aeruginosa, and this was confirmed by plate assay was processed as per the protocol detailed in Venkatakrishnan et al
29
.The sample
29
. Briefly, the sputum
sample was reduced and alkylated with 10mM DTT and 25mM iodoacetamide respectively. Intact cells, cell debris and insoluble mucins were removed by centrifugation for 30 min at 10,000 rpm. Protein concentration was determined using Direct Detect Spectrometer (Merck Millipore, USA). AcroWellTM 96-well plates with BioTraceTM PVDF membrane (Pall Corporation, USA) were activated and washed with methanol and PBS respectively followed by saturation of the membrane with pre-treated sputum. Unbound sputum was decanted and the membrane was washed thrice with PBS. Corresponding flagella (fliC, fliD, flgE, , flgK, flgL, and fliF (encoding MS ring protein)) and pili (pilJ) gene transposon mutants of P. aeruginosa PA14
8 ACS Paragon Plus Environment
Page 9 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
strain
30
along with PAO1 and PASS1-3 strains were used for binding assays. PA14, PAO1
and CF strains PASS1-3 were cultured out of frozen stock on LB agar plates with minimal passaging. Individual colonies of PA14 mutants were inoculated into LB broth and grown overnight. PASS strains and PAO1 were inoculated into SCFM medium and grown till mid-log phase. Both PA14 and PASS1-3 strains were treated independently and cells were harvested by centrifugation and washed twice with PBS, labelled with SYBR Green dye. Cell density was adjusted based on the OD600 of cultures. An equal number of labelled cells were added to the pre-treated membrane in triplicates and incubated for 1h with constant shaking at 300 rpm. Unbound cells were removed by washing with PBS thrice. Fluorescence intensity was measured at 458nm (excitation) and 520nm (emission) and normalized to background fluorescence of the bacterium and sputum alone. Fluorescence Intensity was used as a measure of the binding capacity of the respective strain to the sputum. Plate motility assay: PASS1-3 strains were inoculated in SCFM and grown overnight. The cells were collected by centrifugation, resuspended in SCFM and density was normalized based on OD600.. An equal volume of each density-adjusted cell suspension was spot plated on SCFM (0.4% agarose) semisolid plates, in biological triplicates, and incubated at 37°C for 12 hours. Plates were imaged and sizes of the colonies were used as the measure of motility. Determination of minimum inhibitory concentrations (MIC) for antibiotics: Minimum inhibitory concentration (MIC) of a panel of antibiotics for PASS1-3 strains and PAO1 was determined. We utilized three antibiotics (moxifloxacin, polymyxin and tobramycin) from a selection of antibiotics used in CF patient treatment (Supplementary Table1). Colonies of PASS1-3 and PAO1 strains were inoculated and grown in an overnight culture in SCFM till the mid-log phase. Cells suspension turbidity was adjusted to OD600= 0.7. An equal number of cells were inoculated into 96-well plates containing serially diluted antibiotics prepared in SCFM in triplicates. A drug-free SCFM medium blank control was also included. Plates were 9 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 36
incubated on an orbital shaker at 37°C till the mid-log phase. Growth was measured at 600nm and compared with the blank control.
10 ACS Paragon Plus Environment
Page 11 of 36
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Results and Discussion: Experimental overview: The aim of this work was to characterize the membrane sub-proteomes of P. aeruginosa strains isolated from sputum of CF patients grown in SCFM and M9-glucose medium. We used iTRAQ-MS to compare the membrane protein profiles of three clinical CF strains of P. aeruginosa (PASS1, PASS2 and PASS3)
11
to the well-characterised laboratory/reference
strain, PAO1 (Supplementary Table 1). Previously, Palmer et al
12
demonstrated that the
growth rates, nutritional preferences and gene expression profiles of P. aeruginosa, when grown in SCFM and CF patient’s sputum are similar. Hence, we rationalised that growth in SCFM would mimic the nutritional conditions of CF lungs and induce expression of different membrane proteins compared to M9-glucose minimal medium. As the strains used in this study were obtained from CF patients of different age groups, disease severity and exposure to antibiotic medications (Supplementary Table 1) they illustrate some of the diverse genotypic and phenotypic characteristics of P. aeruginosa observed clinically. Our study is unique since all previously reported P. aeruginosa membrane proteome studies
6-10
used model strains
such as the wound isolate PAO1, which does not necessarily represent the genotypes and conditions needed for adaption to the CF lungs 31. The experimental design used in this study is outlined in Figure 1A, B. Bacterial cells were collected at mid-logarithmic growth (Supplementary Figure S1), lysed, then enriched for membrane proteins after treating with sodium carbonate prior to ultracentrifugation
14
.We
utilized 2D-LC/MS-MS coupled with iTRAQ 4-plex labelling for comparative quantitation. Six independent experiments representing three biological replicates of cells grown in the two different media were performed. Bioinformatics characterization of identified proteins: Mass spectrometry profiling of P. aeruginosa strains revealed 3171 and 2442 proteins (Global protein FDR